Compare traditional AM to theoretical DSB for QRP?

[VE3GZB]

It's just an educated guess on my part but wouldn't DSB offer a greater "focus" of RF energy over traditional AM (by eliminating the carrier and transmitting only the sidebands, you could "spend more power" where the intelligence matters)?

(SSB is too much of a PITA to generate, with the filtering, etc....so why not try DSB)

I'm thinking while the weather is still so cold and I can't play in the garage on building a modulator of cobbling together a QRP DSB rig using old Cossor tubes and I'm wondering what thoughts there are on employing DSB on QRP (I like the idea of using old old tubes and QRP to make distant contacts)?

I tried QRP AM with these old tubes (my boss got them for me the Xmas before last), but 1 watt AM just goes nowhere. So I was wondering if I took the oscillator heart of this QRP AM Heising rig I assembled in the fall (see attachment) and used a few more tubes, created a balanced modulator/power amplifier, fed it from the oscillator and tried to make it work?

73s
geo
VE3GZB

[KX5JT]

QRP AM will be an excersize in patience and frustration.  DSB will be much like SSB in that a BFO or product detector is still needed on the receiving end, thus your contacts will be with SSB folks who may complain about your huge bandwidth. (Well the stations on nearby frequencies might).  You'll more likely work the SSB qrp frequencies and will not likely be any sort of force around the usual AM frequencies.

AM.  All the sugar and calories!  Carrier and power enough to quiet the static and enjoy the nice audio.  AM is not really about QRP Geo.

Just my opinion.

[AB2EZ]

Geo

This might make an interesting project for you... and it might make for some interesting experiments on the air. If conditions are good, and if you have a decent antenna, you will have enough power at the receiving end to communicate. I.e. a 1 watt signal is 20 dB down from a 100 watt signal... and their are plenty of situations where the propagation conditions make received signals 20dB larger than the minimum required for communication. [Keep in mind that mobile stations operating on bands like 75 meters radiate at most a few percent of their transmitted power... the rest going into non-radiating "evanescent" modes that are absorbed by ground losses and nearby objects*]

The biggest obstacles will be interference from stronger signals at the location of your intended receiver, and unwillingness of people who are used to strong signal operation to put up with the difficulties of copying your signal in the presence of fading and interference (and, in some cases, their local noise levels).

Note that there is an increasing number of AM operators who have software-defined-radio receivers; and which can employ synchronous AM demodulation. Therefore, by using a (nearly) balanced modulator, and transmitting a modest amount of carrier (e.g. 10% of full carrier), you can move most of your total power into the sidebands, while still sounding as if you have full carrier.

*In my experience, there are two, opposing points of view on this issue of radiation efficiency. There are people who understand propagation theory, and who "violently agree" that a short antenna radiates most of its power into lossy, non-propagating, evanescent modes (sometimes referred to as the "near field", and sometimes captured in the concept of "ground resistance v. radiation resistance"). Those people will say "Of course, everyone knows that a short antenna has a low radiation efficiency... regardless of how it is constructed (loading coils etc.).

On the other side, there a people who "violently agree" that you can make a short antenna with a high radiation efficiency if you use properly designed loading coils, uning capacitors etc.

For myself, I know that, regardless of the design of the antenna (loading coils, etc.) a short antenna will radiate most of its power into non propagating "evanescent" modes, and only a small fraction of its power into propagating modes. Thus, a short, mobile antenna (operating on 75 meters) will radiate only a few percent of its power into propagating modes that can reach a distant receiver. The maximum radiation efficiency is, roughly, (l/d) x (l/d), where l is actual length of the antenna, and d is a half wavelength (for a doublet antenna) or a quarter of a wavelength for a vertical antenna over a good ground.

Best regards
Stu

[k4kyv]

Note that their is an increasing number of AM operators who have software-defined-radio receivers which can employ synchronous AM demodulation. Therefore, by using a (nearly) balanced modulator, and transmitting a modest amount of carrier (e.g. 10% of full carrier), you can move most of your total power into the sidebands, while still sounding as if you have full carrier.

That will work.  With a synch detector, if the local oscillator has something to lock onto, the demodulated signal will sound identical to full carrier AM. And with DSB you won't have a sideband filter in line to pinch your audio.

You need the pilot carrier to give the synch detector a reference point.  Otherwise the receiving station must narrow up the selectivity so as to receive in SSB mode, since it is impossible to manually adjust the BFO exactly on frequency and exactly in phase, which is a requirement for proper DSB reception.

This is where the ESSB crowd is missing the boat.  If they would transmit a pilot carrier, about 20 dB down from the peak output, the other station  could use a synch detector and completely eliminate the frequency error.  I have used my synch detector and carefully tuned in some of the ESSB stations (ones who are doing it right) with the BFO in manual mode, and the reception was near broadcast quality. The synch detector works equally well with DSB and SSB if it has some reference for the BFO to lock onto.

[VE3GZB]

The biggest obstacles will be interference from stronger signals at the location of your intended receiver, and unwillingness of people who are used to strong signal operation to put up with the difficulties of copying your signal in the presence of fading and interference (and, in some cases, their local noise levels).

Yes, I've been noticing this for quite some time....it seems in my opinion that as technology has improved and provided more to people in general, people in return demand ever more and give back ever less (present company excepted of course). George Carlin, were he alive today, would definitely make a good skit about it.

Thanks for the input! If I come up with something, I'll post about it here with pics.

73s
geo
VE3GZB

[flintstone mop]

Hi Don
 I don't want to go way off topic, but where is a reliable reference for ESSB set-up and use?

Fred

[k4kyv]

Hi Don
 I don't want to go way off topic, but where is a reliable reference for ESSB set-up and use?

Fred

http://www.nu9n.com/home.html

[WU2D]

Geo and Don,

Geez my Marauder does that 20 or 30 db down SSB right now - the carrier null parts need a look and a few need to be replaced like the scratchy pot!

Here are a few fun, old and mostly obsolete symbols that represent AM. There are more weird ones too.

A3: AM Full Carrier with both sidebands ( A- 3- None is AM as we know it)

A3A: mode of single sideband with -16dB pilot carrier (this and A3H and A3R are the common narrowband "AM" used in most commercial AM communications gear)

A3B: Independent Sideband (ISB- reduced carrier plus two sidebands containing different signals)

A3C: Vestigial sideband (e.g. NTSC or PAL)

A3D: AM Pulse Modulation

A3E - Telephony using amplitude modulation: double sideband.

A3H: AME or AM compatible (USB with full carrier or only slightly reduced by 3-6 dB with only upper sideband)

A3J: telephony; single sideband with suppressed carrier

A3R: Single-sideband with reduced or variable carrier

[w1vtp]

Couldn't my CE 20A do the DSB suppressed carrier?  It seems to me I could null the carrier and just unbalance the null a tad and that would give the necessary "pilot carrier" for good sync lock.

Al

[W6TJK]

Hmm,
sounds like a KWS1 in AM mode to me.

the idea of injecting (leaking ?) the carrier back onto a SSB signal is interesting.

Just get the audio BW up over 4kHz and folks will be happy.

Synch AM is very very nice in these nasty, noisy band condx.

t

[N2DTS]

There seems to be no advantage of dsb with carrier over ssb with carrier (KWS-1).
You need a wider filter for dsb, more noise.

Modern ssb rigs can sound very good to me, they seem to be right on freq, no guesswork in tuning them in.
The flex is set for 500 Hz jumps, so when I click on a signal, there is no fine tuning needed for 99% of signals I hear.
With a tuning knob, I could never tune things in as I have zero pitch, or whatever it takes to tell if a signal sounds natural...
When people run old rigs, I tune and tune and cant decide what sounds right. Then someone comes back to the guy slightly off freq, and its tune tune tune...

But with modern rigs, it seems there can be 6 guys in a round table, all are exactly on freq, and they all sound normal. There seems no need for a carrier...
If you turn the rf gain down a bit, the noise between signals is low.

Still, I like the carrier and two sidebands, partly so you can tell the other guy is still sending...


Brett

[w1vtp]

There seems to be no advantage of dsb with carrier over ssb with carrier (KWS-1).<snip>...

Brett

Probably not, Brent.  I hardly ever listen to both sidebands  on AM anyway.  It would just be the fun of finding yet another way of making the old and the new stuff work together.  I wonder if a CE 20a could be deyellafied - on Am at least.  The AF phase network would probably limit SB rejection anyway but who would care about that on AM

Al

[Steve - WB3HUZ]

CW would be even better, but this is the AM Forum.   

[k4kyv]

Here are a few fun, old and mostly obsolete symbols that represent AM. There are more weird ones too.

A3: AM Full Carrier with both sidebands ( A- 3- None is AM as we know it)

A3A: mode of single sideband with -16dB pilot carrier (this and A3H and A3R are the common narrowband "AM" used in most commercial AM communications gear)

A3B: Independent Sideband (ISB- reduced carrier plus two sidebands containing different signals)

A3C: Vestigial sideband (e.g. NTSC or PAL)

A3D: AM Pulse Modulation

A3E - Telephony using amplitude modulation: double sideband.

A3H: AME or AM compatible (USB with full carrier or only slightly reduced by 3-6 dB with only upper sideband)

A3J: telephony; single sideband with suppressed carrier

A3R: Single-sideband with reduced or variable carrier

But notice that none of the above specifically cover double-sideband with reduced or suppressed carrier. A3E simply refers to "amplitude modulation, double sideband". So while every conceivable variation of SSB (suppressed, reduced, variable, pilot carrier)  is each listed as a separate emission mode, all variations of double sideband AM are lumped together as the same mode.

Those emission designators were contrived sometime in the 80's.  Johhny Johnston's Private Radio Bureau took advantage of the new emission type chart to make a bogus argument in their fraudulent AM power proceeding, by claiming that AM was "only one of more than 1200 emission types" used in the amateur service, and therefore was so insignificant that it deserved no special treatment, specifically regarding a proposed permanent extension of the grandfather clause. 

There seems to be no advantage of dsb with carrier over ssb with carrier (KWS-1).
You need a wider filter for dsb, more noise...

SSB with carrier is not the same thing as AM.  It is SSB with poor carrier suppression, and inherently produces distortion with an envelope detector, due to the fundamental principles of the modulation process.  SSB with carrier includes phase modulation components, known as "quadrature distortion", which become significant whenever the modulation percentage exceeds about 20% or 30%.  It makes the signal sound like AM with selective fading, or like SSB tuned in on an AM receiver with the BFO turned off. With higher percentages of modulation it must be copied with the BFO turned on, just like a regular SSB signal.

With DSB, the quadrature distortion produced by each sideband is cancelled out by that of the other.  This can be readily demonstrated using vector diagrams.  The phase variations cancel and leave only amplitude variations to register at the envelope detector.

But the audio voltage derived from the two sidebands is additive, so that a DSB signal produces twice the audio voltage at the  detector than would only one sideband, assuming identical amplitude for each sideband in both cases. Twice the voltage means 4 times the power.  So, by chopping away one sideband of a DSB signal you are effectively reducing the sideband power by 6 db, or to 25% of the original.  If you use a wider filter to accommodate both sidebands, this allows 3 dB more noise to enter the passband, so the net difference in signal-to-noise ratio is reduced to 3 dB.

But with modern rigs, it seems there can be 6 guys in a round table, all are exactly on freq, and they all sound normal. There seems no need for a carrier...
If you turn the rf gain down a bit, the noise between signals is low.

Other than the statement about no need for a carrier,this is true, if a synchronous detector is used.  A synch detector is a product detector with a PLL circuit added to lock the BFO exactly on frequency and in phase with the original AM carrier. The original carrier serves only as a reference; it is the local BFO that actually demodulates the signal at the product detector, just as in the case of a modern SSB receiver.  And, with the synch detector, AM with one carrier is also properly demodulated, since the local BFO demodulates it at the product detector.  There is a need for the carrier, but only enough to give the BFO a reference to lock onto.  The carrier can be reduced 16 to 20 dB without the synch detector knowing the difference.  Likewise, one sideband can be lopped off, and the signal will be demodulated as a SSB signal.

Since DSB produces a net increase in s/n ratio, but the wider passband introduces more noise, for a given total sideband power and optimum receiver bandwidth for each, DSB and SSB produce exactly the  same s/n ratio at the detector.  If you run 500 watts (ignoring the carrier) with DSB and 500 watts with SSB, the s/n ratio will be exactly the same at the output of the detector.  That means  the DSB signal would have 250 watts in each sideband while the SSB signal runs 500 watts.  If one sideband is lopped off from the DSB signal, that leaves only 250 watts in the other sideband so naturally, the signal is down 3 dB compared to the 500 watt SSB signal.

The use of the synchronous detector is becoming more significant with the advent of the software defined radios, such as Softrock and Flexradio, since these black boxes include synch detectors by default, so synch detection is becoming more widely available to amateurs.

There is an urban legend that has propagated ham radio circles for decades, that when receiving AM with a diode detector, only one sideband is detected at the receiver because the other sideband is "rectified away".  I have heard both SSB'ers and AM'ers argue that this is true, but nothing could be further from the truth. Of course, the bandwidth of the receiver can be set so narrow that it can pass only one sideband at a time, but this is another issue entirely. This mistaken concept was propagated by QST sometime in the early 50's during the high-pressure pro-SSB propaganda campaign, when the front cover showed the photo of the rectified envelope pattern of an AM signal, implying that when the diode detector rectifies the signal and lops off everything on the opposite side of the baseline, it was lopping off one of the sidebands.

Surprisingly, no-one ever called QST's hand on this, or if they did, they refused to publish it. I always thought they did a disservice to the amateur community by deliberately promoting an incorrect technical concept to the largely non-technical ham community.

[AB2EZ]

Don

Well said!

Stu

[N2DTS]

The flex radios have a dsb button, next time I work someone with a flex, I should try it and see what happens.
I dont know how its supposed to work, does the flex insert a carrier?
How does it know what freq to generate the carrier on, or does it just center it on the listening frequency?

Don,
So that is why the KWS-1 always sounded SO bad in the AM position (although some guys who run them dispute that).

The below quote was about listening to ssb, in the ssb mode, on a modern rig...

Brett


Quote
But with modern rigs, it seems there can be 6 guys in a round table, all are exactly on freq, and they all sound normal. There seems no need for a carrier...
If you turn the rf gain down a bit, the noise between signals is low.

[k4kyv]

The KWS-1 and Drake TR-7 both used SSB+carrier and sounded like crap on an envelope detector.

The ESSB boys are missing the boat by not transmitting a pilot carrier and using a synchronous detector.  That would allow them to approach broadcast quality.

[N2DTS]

Don,
Some of them sound very good to me.
I would go so far as to say some of those guys sound better than I do, and better then 80% of the AM signals I hear.
At least they do on my flex (being very smug here)....

Nick played a recording of me last weekend and I sounded nasty, the new EQ needs adjustment (not so smug).


Their big problem sounding good is the bandwidth limits the radios have built in, and there are some that can be modified to pass 4Kc or maybe more, but most need mods to do so, or cant.

I have no problem with ssb, its just another mode, but never hear much that is interesting on ssb.

On a quiet band, with stronger signals, it can sound almost like AM to me.

Brett

[k4kyv]

The ESSB can sound very good if it is done right.  Some just try to brute-force a wider audio bandwidth through a stock 2.1  or 2.7 kc filter using a lot of equalising.  Like trying to drive a square peg through a round hole. The only thing that will accomplish is distortion and splatter, but if a wider filter is used, or one of the newer technology rigs with digital filtering with variable bandwidth, and good audio is fed into the rig (garbage in = garbage out), the signal can approach broadcast quality.

Where they are missing the boat is by not transmitting a low level pilot carrier for reference and copying with a synch detector.  With total sideband suppression, you need a  good ear to tell when the signal is tuned in exactly, and it helps if you are familiar with the other person's live voice.  With the pilot carrier and PLL, the received signal would be tuned in automatically exactly on the nose, without frequency error, but very little power would consumed by the reduced carrier so the power efficiency would be about the same as that of regular slopbucket.

With all the SDR's out there, this is now feasible.  Previously, only a few hams would shell out hundreds of bucks for a good analogue synchronous detector like the Sherwood.

[Steve - WB3HUZ]

With total sideband suppression, you need a  good ear to tell when the signal is tuned in exactly, and it helps if you are familiar with the other person's live voice.

This is basically not true when the transmit and receive station are both using modern synthesized radios. If the TX station is on 3630 kHz, it's easy for the receive station to tune to 3630 kHz and be within a few Hertz or less of the TX frequency. No special tuning or ears required.

[w1vtp]

With total sideband suppression, you need a  good ear to tell when the signal is tuned in exactly, and it helps if you are familiar with the other person's live voice.

This is basically not true when the transmit and receive station are both using modern synthesized radios. If the TX station is on 3630 kHz, it's easy for the receive station to tune to 3630 kHz and be within a few Hertz or less of the TX frequency. No special tuning or ears required.

I basically agree with you Steve but with my system the Rx AF response goes basically down to close to DC.  In a recent test with Steve QIX, I could see the speaker cone move at the lowest frequencies.  I didn't have the AVC set up correctly so I got some pretty weird sound at the low end (I think I have that corrected now). 

Don's suggestion of a pilot carrier would work especially nice with my system as I would not get that slow cone movement (which is rather disturbing to watch) with the low frequencies because I am now phase coherent with the incoming signal. So I think it's a worthwhile effort to tinker around with a pilot carrier if some of the stations have AF response down to close to DC (around 5 to 10 Hz).

Great thread

Al, Violet Toilet Paper

[Steve - WB3HUZ]

Those low freqs are hard on the speaker, assuming your system can pass them. There's an easy fix though, just cut off everything at 20 Hz or 50 Hz. There are no voice components at these frequencies anyway, so why pass them?

[K5UJ]

I understand Flex radio automatically cuts off tx audio at and below 50 hz in SSB.  doesn't do it in AM.  has to do with meeting specs for carrier and opp. sb suppression I think.

Heard my first carrier + LSB rig a few days ago.  Not sure what it was, TR7 KWS1 or something else but I thought it sounded noticeably worse than regular constant carrier + two sidebands.

For me, I can tell a difference in recovered audio quality between the hi-fi ESSB and AM.  No matter what pains an ESSB guy goes to, his audio will always have a certain subtle characteristic SSB sound.  Some of it has to do with the absence of carrier blanking out the background noise, but only A3 type AM can come close to what I would describe as class A audio amp quality.

In my salad audio days doing, or rather attempting to do, the ESSB hi fi audio thing, I pumped up my low and high frequencies and attenuated the mid-range.  Some of the ESSB guys do this and now I find it annoying.  Operators who possess naturally bassy voices don't need it and those who don't sound a little creepy with it.  It makes them difficult to copy in poor condx so I wind up narrowing the low part of the passband to cut off at 300 hz or higher to get intelligible audio.  At the high end I get lispy amounts of consonant essing.   This type of tx audio sounds pretty bad in a standard SSB passband of around 3 KHz or less and it doesn't have to.

Rob

[k4kyv]

If the TX station is on 3630 kHz, it's easy for the receive station to tune to 3630 kHz and be within a few Hertz or less of the TX frequency. No special tuning or ears required.

This is where the GPS reference lock signal would be useful.

[N2DTS]

Don,
It seems to me, they have it covered from a frequency standpoint, modern radios might be 1 or 2 Hz off, which is not noticeable on ssb.
The flex is more accurate than the freq counter I have....

It does not sound as good as the best AM, but I am sure there are ssb guys who sound a lot better than I do on AM.

Brett